摘要
Echogenic fetal lung lesions form a heterogeneous group that display a variety of features. The difficulty in distinguishing and defining them is reflected in the number of papers debating the pertinence of the various classifications published in the past 20 years. As pointed out by Langston: ‘Ultrasound was not in use when the currently used definitions and classifications of congenital malformation of the lung were elaborated.’1. Stocker's original classification of congenital cystic adenomatoid malformation (CCAM) in the 1970s2 was based on postoperative or postmortem pathological data. When prenatal diagnosis of CCAM became available using ultrasound, Stocker's classification was applied, despite the inherent differences pre- and postnatally. This classification was modified in 20013, now being referred to as congenital pulmonary airway malformation (CPAM), and comprising five types of lesion based on the site of origin of the malformation (trachea, bronchus, bronchiole, alveolar duct or distal acinar). This new classification is even less applicable to prenatal diagnosis, particularly as not all types are cystic or adenomatoid on pathological examination; pathological findings, especially histology, cannot always be related to prenatal ultrasound findings. In order to reclassify echogenic fetal lung lesions, Achiron et al.4, developed a new classification in 2004 which enables separate delineation of each fetal lung component involved in the anomaly: parenchyma, arterial supply and venous drainage. It is based on prenatal ultrasound and Doppler findings rather than on pathological ones. Although an interesting concept, this classification is unlikely to be widely adopted because it has the potential to cause even more confusion in this already obscure field. Their classification involves five types of lesion, which they called ‘dysplasia’. As mentioned by Sebire5, the term ‘dysplasia’ is inappropriate because ‘lung dysplasia’ is already used for congenital lung pathologies which are not malformations but abnormal development of the lung without the presence of a mass lesion. Type 1 is already termed ‘pulmonary agenesis’ and the fifth ‘miscellaneous’ category includes pathologies with completely different embryological origin. In this issue of the Journal, Cavoretto et al.6 report the prenatal diagnosis and outcome of echogenic fetal lung lesions. They have collected an impressive number of cases of CCAM and pulmonary sequestration (PS), but also studied fetuses with congenital high airway obstruction syndrome (CHAOS). The latter is a well-defined clinical entity and does not raise the same issues as do PS and CCAM, because the diagnosis is straightforward and there is no differential diagnosis. Cavoretto et al. stated that most existing studies on hyperechogenic lung lesions reported fewer than 30 cases, making it impossible to provide definitive conclusions as to the natural history and prognosis. Unfortunately, although being well-conducted and reporting a high number of cases, their study might also fall short of fully addressing these issues. Many pathological entities are characterized by echogenic fetal lung lesions: CCAM and PS in particular, but also bronchial atresia and congenital emphysema. Bronchial atresia can be diagnosed prenatally when it is proximal, while distal lesions are more often diagnosed during childhood7. Congenital emphysema can mimic Type III CCAM, having disappeared by the time of the third-trimester scan but being present on postnatal tomodensitometry8. In the study of Cavoretto et al.6, 154 cases with CCAM survived, of which 83 did not need surgery, meaning that pathology results were not available in half of the cases. Detection of the lesion also depends on the time at which X-ray and tomodensitometry are performed after delivery; if they are performed too early, when no clinical sign of respiratory distress is present, the lesion can be missed. Therefore, some cases may have been misdiagnosed as CCAM and may in fact have been PS, bronchial atresia, emphysema or even (as mentioned by the authors) transient bronchial tree obstruction due to mucus plug. As pointed out by Cavoretto et al.6 and many others, mixed entities have been reported, which make things even more complicated for fetal medicine specialists9, 10. Pathologists have been fighting for years in order to classify congenital lung lesions1, 11, 12; prenatal specialists would be well-advised to simplify the matter and this is what Cavoretto et al.6 have succeeded in doing. Prenatal classification should: (i) differentiate large from small cysts (which could be classified into one of three sub-types as suggested by Adzick a long time ago13) because their in-utero therapeutic options will differ, and (ii) identify the presence of a systemic feeding vessel. Whether prenatally we call them CCAM or sequestration might be, in fact, irrelevant. Particular attention should be paid to distal bronchial atresia, whose prognosis is poor because, although the fetus will not always develop hydrops, it is difficult to treat postnatally and most often is associated with contralateral lung hypoplasia. This diagnosis should be made in the presence of a persistent unilateral hyperechogenic lung with a small hypoechogenic cyst corresponding to the proximal dilated portion of the bronchus. Ultimately, however, does it really matter whether we are able to differentiate prenatally a CCAM from congenital emphysema or PS? It all depends on our aim. From a fetal management point of view and for the evaluation of neonatal prognosis, such differentiation is inconsequential. Cavoretto et al.6 have confirmed what was already known by fetal medicine specialists, that the appearance of hydrops marks a turning point in the course of such diseases and considerably affects prognosis. However, it should be kept in mind that, even in the presence of hydrops, spontaneous regression of the lesion is possible and that in some cases drainage of a cyst or of hydrothorax or coagulation of a feeding vessel is feasible. PS is often associated with other malformations, and particular attention should be paid to this risk when a systemic vessel is visualized. Neonatal surveillance will be similar and should be carefully conducted for all lesions depending on the size of the lesion and its repercussions on homeostasis. Accurate diagnosis will be made postnatally if surgery and pathological examination are performed. If no surgery is needed the prognosis is usually good but the diagnosis will remain unconfirmed. From the perspective of natural history, even with this large series of Cavoretto et al.6, there remain unanswered questions. The embryogenesis and pathogenesis of fetal lung lesions are still unclear and many hypotheses have been suggested. Some authors suggest that bronchial atresia is the primary defect leading to the development of CCAM1, 14, 15. Langston1 suggested the possibility of a malformation sequence based on the level, timing and degree of airway obstruction during development as a unifying pathogenetic mechanism for the various developmental lung lesions. Kunisaki et al.16 even suggest that during pathological examination particular attention should be paid (microdissection with bronchography) to the visualization of atresia because, in their series, the prevalence of atresia was high, having been detected by pathologists trained for this examination. Shrinkage or disappearance of echogenic lesions is still an unresolved issue. Many hypotheses have been suggested to explain this, such as delayed maturation of pathological tissue17-19, spontaneous occlusion of a feeding vessel4 (that might not have been seen on Doppler examination), transient bronchial occlusion20 and dysregulation in the balance between the natural processes of apoptosis and proliferation21. However, confirmation of these hypotheses will be very difficult to obtain. We now have sophisticated ultrasound equipment with Doppler capability; it should therefore be possible to improve the prenatal evaluation of these lesions, especially through vascularization studies. A prospective study should be performed comparing prenatal images with pathological examinations, looking specifically for bronchial atresia. Molecular biology studies of pathological specimens, investigating different factors involved in these diseases, might give clues to end the controversies surrounding histological classification and the mechanisms involved in the disappearance of echogenic lung lesions.